fix: normalize null tool input from Anthropic adapter#266
Merged
AlemTuzlak merged 5 commits intoTanStack:mainfrom Feb 7, 2026
Merged
fix: normalize null tool input from Anthropic adapter#266AlemTuzlak merged 5 commits intoTanStack:mainfrom
AlemTuzlak merged 5 commits intoTanStack:mainfrom
Conversation
When Claude occasionally produces a tool_use content block with no
input_json_delta events (or with "null" as the partial JSON),
JSON.parse("null") returns JavaScript null — which passes the truthy
check but isn't {}. This causes downstream Zod schema validation to
fail, and the agent loop stalls silently.
Changes:
- Anthropic adapter (content_block_stop): normalize parsed input to {}
when JSON.parse produces null or a non-object value
- Anthropic adapter (formatMessages): same normalization when
reconstructing tool calls for conversation history
- executeToolCalls + ToolCallManager.executeTools: normalize "null"
argument strings to "{}" before parsing
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Contributor
📝 WalkthroughWalkthroughValidated and normalized JSON inputs for tool-handling: the Anthropic text adapter now ensures parsed JSON is an object before assignment, and tool-call handlers trim input, convert the literal "null" and empty/whitespace to "{}", then parse safely to avoid non-object or empty inputs being used. Changes
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~20 minutes Suggested reviewers
Poem
🚥 Pre-merge checks | ✅ 3✅ Passed checks (3 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing touches
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
|
View your CI Pipeline Execution ↗ for commit 0b229d3
☁️ Nx Cloud last updated this comment at |
@tanstack/ai
@tanstack/ai-anthropic
@tanstack/ai-client
@tanstack/ai-devtools-core
@tanstack/ai-gemini
@tanstack/ai-grok
@tanstack/ai-ollama
@tanstack/ai-openai
@tanstack/ai-openrouter
@tanstack/ai-preact
@tanstack/ai-react
@tanstack/ai-react-ui
@tanstack/ai-solid
@tanstack/ai-solid-ui
@tanstack/ai-svelte
@tanstack/ai-vue
@tanstack/ai-vue-ui
@tanstack/preact-ai-devtools
@tanstack/react-ai-devtools
@tanstack/solid-ai-devtools
commit: |
Merged
This was referenced Feb 7, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
#265
Summary
When Claude occasionally produces a
tool_usecontent block with noinput_json_deltaevents (or with"null"as the partial JSON), the Anthropic adapter'scontent_block_stophandler passesnullas the parsed input instead of defaulting to{}.This causes
executeToolCallsto fail Zod schema validation (sincenullisn't an object). The error result is correctly captured, but the agent loop stalls after the failed tool execution — the model either doesn't produce a follow-up text response, or the follow-up text isn't emitted through the stream.The end result for users: the assistant says "Let me search for that..." then silence. No error, no follow-up.
Root Cause
In
@tanstack/ai-anthropic—adapters/text.tsThe
processAnthropicStreammethod, inside thecontent_block_stophandler:When
existing.inputis the string"null"(from the model streamingpartial_json: "null"),JSON.parse("null")returns JavaScriptnull— which passes the truthy check but isn't{}. The resultingTOOL_CALL_ENDevent hasinput: null.In
@tanstack/ai—tools/tool-calls.tsDownstream in
executeToolCalls:toolCall.function.argumentsis"null"(set bycompleteToolCallviaJSON.stringify(null))argsStr = "null".trim() || "{}"→"null"(truthy, so no fallback)JSON.parse("null")→nullparseWithStandardSchema(tool.inputSchema, null)→ Zod throws (null is not an object)Changes
Anthropic adapter (
content_block_stop): Normalize parsed input to{}whenJSON.parseproducesnullor any non-object value:Anthropic adapter (
formatMessages): Same normalization when reconstructing tool calls for conversation history (preventsinput: nullbeing sent back to the Anthropic API on subsequent turns).executeToolCalls+ToolCallManager.executeTools: Normalize"null"argument strings to"{}"before JSON parsing — a defense-in-depth fix for the same issue at the tool execution layer.How to Reproduce
This is intermittent and depends on model behavior. We've observed it with
claude-sonnet-4-5when the model generates atool_useblock with no meaningful arguments (e.g., after a first tool call that returned zero results). The model seems to "start" a follow-up tool call but doesn't commit to arguments.Typical pattern:
tool_use(with valid arguments) + secondtool_use(with null/empty arguments)Example Session Data
{ "role": "assistant", "content": "Let me search for disc mowers near you in Saskatoon:", "toolCalls": [ { "id": "toolu_1770336584303_0", "type": "function", "function": { "name": "search_equipment", "arguments": "null" } } ] }Test Plan
JSON.parse("null")→ normalizes to{}existing.input→ defaults to{}Summary by CodeRabbit